Dataset card: title, logo, links, viewer off

README.md

@@ -1,5 +1,6 @@
 ---
 pretty_name: pyrregular
+viewer: false
 license:
 - cc-by-4.0
 - cc-by-nc-sa-4.0
@@ -11,10 +12,25 @@ license:
 license_name: dataset-specific
 ---
 
-# pyrregular datasets
+![Pyrregular: irregular time series datasets and benchmarks](https://github.com/fspinna/pyrregular/blob/main/assets/images/logo_01.png?raw=true)
 
-Irregular time series datasets converted to the format of the
-[pyrregular](https://github.com/fspinna/pyrregular) library ([paper](https://openreview.net/forum?id=qetBM8nLkf)).
+# Pyrregular: Irregular Time Series Datasets and Benchmarks
+
+**Published at ICLR 2026** · 38 datasets, including the benchmark of
+34 datasets and 12 classifiers ·
+[GitHub](https://github.com/fspinna/pyrregular) · [Documentation](https://fspinna.github.io/pyrregular/) ·
+[PyPI](https://pypi.org/project/pyrregular/) ·
+[Paper](https://openreview.net/forum?id=qetBM8nLkf)
+
+Naturally irregular time series (uneven sampling, missing observations, signals
+recorded at different times, variable-length sequences) in one standardized
+format, for classification and regression. These are the datasets of the
+[pyrregular](https://github.com/fspinna/pyrregular) Python library, including the 34 of the ICLR 2026
+benchmark. The files are HDF5, read by the library (the dataset viewer is off):
+
+```bash
+pip install pyrregular
+```
 
 ```python
 from pyrregular import load_dataset